UBC Faculty Research and Publications

The Development of Intelligent Systems to Support Older Adults and Aging-in-Place 2009

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
AlexM_UBC_ipod.mp4
AlexM_UBC_ipod.mp4 [ 92.39MB ]
AlexM_UBC_slides.pdf [ 6.81MB ]
AlexM_UBC_slides.pdf
Metadata
JSON: 1.0084645.json
JSON-LD: 1.0084645+ld.json
RDF/XML (Pretty): 1.0084645.xml
RDF/JSON: 1.0084645+rdf.json
Turtle: 1.0084645+rdf-turtle.txt
N-Triples: 1.0084645+rdf-ntriples.txt
Citation
1.0084645.ris

Full Text

The Development of Intelligent Systems to Support Older Adults and Aging-in-Place Alex Mihailidis, Ph.D P.Eng. Department of Occupational Science & Occupational Therapy 2The Problem  Globally we are faced with the challenge of caring for an increasing number of older adults.  Many individuals are living with diseases and impairments in addition to the normal aging process, for example dementia. Furthermore...  There is a growing shortage of clinicians, nurses, and caregivers.  In combination with the growing number of older adults, we are seeing  a drastic increase in workload and burden.  This not only true in institutions but in the home as well! 3 4The Result  There is an increased need for new treatment options and solutions.  Solutions need to be robust and easily scalable to meet the needs of this growing population.  These solutions need to help support aging-in-place. 5Overall Vision  Our goal is to develop an intelligent home that can enable aging-in-place, no matter the context or activity. 6Alex, don’t forget to use the soap. You have fallen. Need Help? It’s time for your medication. Are you feeling okay? 7Our Design Philosophy  Develop for real-world contexts, using real- life problems and motivations.  Involve the user from the start to the finish of the design process.  Test new technologies as often as possible throughout the design process.  Universal Design principles to accommodate all potential users (including caregivers). 8Zero-Effort Technologies  Development and testing of intelligent systems that are able to: • Automatically learn about the user and environment • Continuously collect data and information • Use contextual information to autonomously operate and make decisions  This is all accomplished without any input from the user or caregiver. 9Data Sharing The COACH (ADL Cueing) HELPER (ERS) Anti-collision Wheelchair Robotic Stroke Rehabilitation Nursing PDAs Balance Assess. Examples of Our Research 10 Data Sharing The COACH (ADL Cueing) HELPER (ERS) Anti-collision Wheelchair Robotic Stroke Rehabilitation Nursing PDAs Balance Assess. Examples of Our Research 11 The COACH An intelligent cognitive device that tracks a user through an ADL, providing cues when necessary Cognitive Orthosis for Assisting Activities in the Home The COACH 12 Alzheimer’s Disease  The number of people with AD worldwide is expected to grow from 18 million to 24 million by 2050.  There is approximately one new case of AD every 7 seconds!  70 percent of people with Alzheimer’s and other dementias live at home. 13 Alzheimer’s Disease  AD impairs explicit memory, resulting in difficulties in completing activities of daily living (ADL).  The current solution is for a caregiver to constantly accompany the person and provide prompts, support, and monitoring.  This is a very difficult and frustrating experience. 14 The COACH Digital Video Camera Flat Screen Monitor & Speakers 15 System Overview 16 Tracking 17 Planning  The system’s belief monitoring and policy systems use AI to model the handwashing process and actions that may be taken.  Represented using: • Level of impairment • Awareness • Responsiveness 18 Handwashing - Actions  Do nothing: system waits  Call caregiver: system calls for single step assistance  Prompts: • Audio/video • Male voice • Preceded with reminder: “you’re washing your hands” • 3 levels of specificity 19 Prompt Specificity  Minimal “Turn on the water”  Moderate “Ed, pull up on the silver lever in front of you to turn the water on”  Maximum “Ed, pull up on the silver lever in front of you to turn the water on” + 20 Example of Use Future Work  COACH@Home System • Multiple tasks • Easy installation • Aesthetics  Intelligent dialogue management • Speech recognition • Multi-language, accents, vocal invariants 21 22 Examples of Our Research Data Sharing The COACH (ADL Cueing) HELPER (ERS) Anti-collision Wheelchair Robotic Stroke Rehabilitation Nursing PDAs Balance Assess. 23 Data Sharing The COACH (ADL Cueing) HELPER (ERS) Anti-collision Wheelchair Robotic Stroke Rehabilitation Nursing PDAs Balance Assess. Examples of Our Research 24 The HELPER    An intelligent hands- free personal emergency response system to improve safety in the home 25 Overview of Falls  Falls is one of the leading causes of morbidity and mortality in the elderly.  One in three community-dwelling older adults experience at least one fall over the span of one year.  One third of these falls occur in the home. 26 System Components Camera Mic array & speakers 27 System Schematic 28 Fall Detection 29 Dialogue Manager  Once an adverse event has been detected the system must respond appropriately.  Prompting and speech recognition will be used to have a dialogue with the user.  Based on this dialogue, the system will determine the best course of action. 30 Response Actions  Actions that can currently be taken by the system include: • Call a neighbor • Call a family member • Call an operator (e.g. Lifeline) • Call an emergency service (e.g. ambulance)  Future actions may include prompting and reminders. 31 Example of Use Current / Future Research  Focusing on the speech recognition / response system.  Developing a speech database of older adult voices (healthy and not).  Development of an adaptive dialogue that learns the preferences and needs of specific users. 32 Design Approach Revisited  We have learned that applying our design philosophies and approach are extremely important in our successes.  However, involving users has proven to be very costly and time consuming.  In response, we are looking at new tools that we can use to make this approach more efficient. 33 Actors in Design  We are starting to explore the use of actors in the design process.  These actors would simulate different users so that we can test our systems more often during the design process.  Building upon work by University of Dundee and classical use of patient simulation programs by medical students. 34 Actor simulations - Objectives  The objectives of this pilot research were to investigate the following questions: • Can actors believably simulate older adults with dementia? • Can actors simulating older adults with dementia be used to optimize technologies before they go to clinical trials? 35 Actor simulations – Method  Actors were trained using video footage from previous handwashing trials.  Videos were recorded of older adults washing their hands: • 6 were actors portraying dementia (“simulated”) • 6 were people with dementia (“real”)  20 videos per person, for a total of 240 videos. 36 Actor simulations – Method  Task 1: Participant randomly shown 20 segments (10 simulated and 10 real) and asked to rate on a scale from 1 (not at all) to 10 (very much): “Was the client behaving in a way a person with dementia would?”  Task 2: Participant shown 10 video pairs containing a simulated and a real trial and asked to choose which video was the actor. 37 Which one is the actor? 38 Actor simulations – Task 1 Av er ag e sc or e (b et w ee n 1 an d 10 ) Real / simulated dementia pair 39 Actor simulations – Task 2  65 % of the time the actor was correctly identified.  “When making my choices, I felt…” • Very sure (0) • Somewhat sure (3) • Somewhat unsure (7) • Very unsure (1) 40 43 Conclusions  Non-traditional tools and techniques, such as AI, have the potential to make environments more usable and safe.  Intended user must be kept in mind and involved often for successful outcomes.  The potential exists to use actors to improve design process.  In addition to technological challenges, development must focus on the social and ethical implications. 42 Acknowledgements  Research has been supported through generous research grants and industrial support from: 43 Contact Contact us:       Tel:   +1 (416) 946-8565 Email:   alex.mihailidis@utoronto.ca   Web:  www.iatsl.org

Cite

Citation Scheme:

    

Usage Statistics

Country Views Downloads
United States 10 1
China 5 14
Japan 5 0
City Views Downloads
Tokyo 5 0
Redmond 4 1
Beijing 3 0
Mountain View 3 0
Shenzhen 2 14
Ashburn 2 0
Sunnyvale 1 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}

Share

Share to:

Comment

Related Items